The SK Hynix Nasdaq listing launched on 10 July, with the South Korean memory giant targeting approximately $28.21 billion (43.14 trillion won) through 17.79 million new American Depositary Receipts, priced at a ratio of ten ADRs per one common share. BofA Securities, Citigroup, Goldman Sachs, and J.P. Morgan are serving as global coordinators, joined by nine other financial institutions, according to Yahoo Finance.
The listing puts two of the biggest names in AI chips side by side on US exchanges: Nvidia (NVDA), the dominant accelerated-computing company, and SK Hynix (SKHY), which makes the high-bandwidth memory those chips depend on. Both trade at valuations that look undemanding given their growth rates. They are not, however, equivalent bets.
SK Hynix Nasdaq Listing: What the Money Is For
Proceeds from the SK Hynix Nasdaq listing are earmarked for expanding South Korean manufacturing and acquiring equipment, including ASML’s extreme ultraviolet lithography scanners, Yahoo Finance reported. The capital plans are substantial. The Elec reported that SK Hynix plans to invest 19 trillion won in an advanced packaging facility, designated P&T7, in Cheongju, targeting a cleanroom opening by end-2027, and 5.9 trillion won in an advanced packaging facility in Indiana with a cleanroom opening planned for the second half of 2028.
The company enters this expansion from a strong position. SK Hynix held more than 35 trillion won in net cash at the end of the first quarter of 2026, and has set a medium- to long-term goal of accumulating more than 100 trillion won in net cash, according to The Elec.
The financial backdrop supports that ambition. Tweaktown reported that SK Hynix posted record full-year 2024 revenues of 66.1 trillion won (approximately $46.1 billion), up 102% year-on-year, with operating profits of 23.4 trillion won, driven by high-bandwidth memory (HBM) sales. Revenue growth has continued: Blocks & Files reported that Q2 2025 revenues reached 22.23 trillion won ($16.23 billion), up 35.4% year-on-year, and the company expects to double its 2025 HBM revenues compared with 2024.
SK Hynix commands 57% of global HBM revenue, according to Counterpoint Research data cited by Bloomberg via Yahoo Finance, measured for Q4 2025. The snippet’s description of a “near-60%” share aligns with that figure. Samsung and Micron are competing hard for the remainder, and the cyclicality of the memory market remains the core structural risk: demand cycles have inflicted severe downturns on the sector before.
The forward price-to-earnings ratio sits at roughly five, a deep discount to Nvidia. Analysts cited by the Wall Street Journal, as reported by Yahoo Finance, note that potential inclusion in the Philadelphia Semiconductor Index (a benchmark tracked by passive funds globally) could follow the US listing, and SK Hynix’s valuation gap relative to Micron Technology may narrow over time.
Nvidia: Vera Rubin and the Scale of the Ambition
Nvidia’s investment case rests on a different type of compounding. The Vera Rubin platform has been accelerated to Q3 2026, with seven distinct chips in the family having completed tape-out. Nvidia is targeting approximately $500 billion in GPU-compute revenue by end-2026, excluding China entirely, the company currently counts its China market contribution as zero, given US export restrictions and government procurement bans.
The hardware specifications underline why the platform has attracted that revenue ambition. Nvidia’s newsroom states that the Vera Rubin NVL144 CPX packs 8 exaflops of AI compute, delivers 7.5 times the AI performance of the GB300 NVL72 system, and offers 100TB of fast memory and 1.7 petabytes per second of memory bandwidth in a single rack. The Nvidia developer blog adds that the Rubin GPU delivers 50 PFLOPS of NVFP4 inference performance and carries a 336-billion-transistor design.
Nvidia’s P/E ratio sits below 20 on the snippet’s figures, and the average analyst price target of $302 implies roughly 50% upside from the current share price. The stock is up less than 10% this year, consolidating gains after an extended run. Its CUDA software platform continues to provide a structural moat: an ecosystem of developers coding against it represents switching costs that competitors such as AMD and Broadcom have not yet dissolved.
The principal risks are a slowdown in AI capital expenditure and erosion of GPU market share. Both are real. But Nvidia’s software lock-in, platform breadth (expanding into CPUs, humanoid robotics, and autonomous vehicles) and the sheer scale of the Vera Rubin revenue target make it a more defensible position than a memory supplier operating in a structurally cyclical market.
SK Hynix’s ADR listing is a genuine milestone, and the valuation discount is real. The Philadelphia Semiconductor Index inclusion thesis, if it plays out, could accelerate the gap with Micron. That catalyst is worth watching.
